Any interest in updated SE Alaska RNAV RNP approach procedures (including Juneau)?

[ALF1 15]

Hey guys, just came across this thread and had to resurrect it to thank you guys for this terrific work !

 

Any chance to do some work to lower minimums for rwy 8 ? I often find myself in trouble with the published LOC approach 3200 Ft MDA.

 

I understand AS has a propietary RNP .15 as well for rwy 8.

[Stearmandriver 2,295]

Any chance to do some work to lower minimums for rwy 8 ? I often find myself in trouble with the published LOC approach 3200 Ft MDA.

 

I understand AS has a propietary RNP .15 as well for rwy 8.

 

Oh, this one's easy:  just fly in Delta colors.  Then your simulation will be entirely accurate!   :Big Grin:

 

Seriously, yeah there's an RNP for 08, and yeah it significantly lowers minimums.  And yeah, I've already built it, as well as about a dozen of our other SE Alaska RNPs, including a few departures.  Once you get the hang of this procedure building thing, it becomes like a puzzle... it's an ok way to kill some time in hotels.  But here's the thing: I had an interesting discussion with someone in the company who was in a position to have an "official" stance on this, and his "official" stance was, "it's not technically illegal but there are people who wouldn't be happy about it right now, so you're better off not making anything public".  And so I won't, excepting the older version of the 26 approach which was already on the internet before I built it.  Sorry, I know that makes this thread kind of a tease now... but maybe in the future it will become less of a hot button issue and I can publish them.

 

I've already shared with a couple RW ASA pilots; if any others would like the up-to-date versions of the procedures, feel free to PM me.  I'll ask a quick question to verify you work for ASA and then email them.

I've never had a GPWS warning in JNU on the approach. Something needs tweaking.

 

Oh, and 8/26 is closed! You're supposed to land on taxiway Alpha.

 

Picky, picky.   :wink:   Yeah, the GPWS warnings are something I wasn't able to eliminate without shoving SALMN unrealistically far west, thus screwing up the approach geometry for the lineup turn.  Which as you can see, doesn't work quite as well as real life anyway, so it doesn't need to be made worse.  The NGX doesn't seem to track an LNAV course quite as precisely as the real plane; it swings wide in turns for some reason, even when allowed to lead the turn (by using flyby waypoints).   This can be compensated for somewhat by moving the WP the turn initiates at back a bit, but I never could get it perfect.  

 

As far as the GPWS, I'm not sure if it's because my procedure gets closer to the hills than real life (looks about right out the window), or if the EGPWS logic in the NGX is a little different than real life, or a combination of the two.  I decided to live with it... because that's the best I could do lol.

 

Oh, and since it took me 2 months to see this response, 8/26 is back open... landing on alpha frowned upon.   :wink:

[ALF1 15]

Doggone it Andrew great job !! I just flew the RNAV 26 approach and what a beauty it is. You have some real nice ground reference points  (the highway) and there are a bunch of trees in the middle of the prairie on the captains side window that can be used as reference to fly over on the turn to final. You do need the Orbx Southeastern Alaska region and the Juneau airport and I don't know if those trees are for real but it helps in the visual turn to final.

 

Navigraph data has TRDWL, SALMN,  and LEMNN. I had been using those waypoints with the headings to do a pseudo approach but his takes it further out many thanks.

 

I will take the opportunity to tell you I admire the airline you work for, have flown in it many times and great service. Gotta love those RNP approaches. And I fully understand the jealousy guarding the unique work and design in these approaches.

[Spin737 102]

 

 

Oh, and since it took me 2 months to see this response, 8/26 is back open... landing on alpha frowned upon. :wink:

 

Good thing I read this. I'm taking 67 up tomorrow. Boy, would I have been embarassed!

 

Well, looks like tomorrow might be an RNP type of day. Looks like Fall has returned to the SE.

[Copper. 78]

https://youtu.be/y0rqln3Dwg0

 

I made a fictional RNP approach to RWY08 (WIP) but I can't get the FMC to generate a Glide Path to comfortably clear Mendenhall Peninsula. I am able to clear the Peninsula by 200 feet continuously, but I don't think it's enough.

 

If I look out to the left whilst passing directly over the peninsula along the RW08 centreline, I see trees and terrain at my level. If I left the lateral protected area (0.2nm either side using RNP 0.10), I'd probably end up crashing. If I miss-set QNH (like I did in the video) I'd probably end up crashing.

 

Making a waypoint closer to the runway didn't seem to help. Anyone able to offer advice? I've added the bits I created for the PAJN.txt SIDSTAR file. Lateral and Vertical obstacle clearance isn't guaranteed as I don't know how to check easily.

 

FIX JN004 LATLON N 58 21.64000 W 135 19.87000
FIX JN003 LATLON N 58 21.49000 W 135 10.18000
FIX JN002 LATLON N 58 23.59000 W 134 50.68000
FIX JN001 LATLON N 58 22.80000 W 134 45.26000

 

APPROACH RNV08 FIX JN002 AT OR ABOVE 2100 FIX JN001 AT OR ABOVE 2100 RNW 08 FIX LEMNN FIX TRDWL FIX MARMN FIX GLAZZ HOLD AT FIX GLAZZ LEFT TURN INBOUNDCOURSE 125 ALT 7000 LEGTIME 1
TRANSITION CUSHI FIX CUSHI AT OR ABOVE 7000 FIX JN004 AT OR ABOVE 6400 FIX JN003 AT OR ABOVE 5200

[Spin737 102]

https://youtu.be/y0rqln3Dwg0

 

I made a fictional RNP approach to RWY08 (WIP) but I can't get the FMC to generate a Glide Path to comfortably clear Mendenhall Peninsula. I am able to clear the Peninsula by 200 feet continuously, but I don't think it's enough.

 

If I look out to the left whilst passing directly over the peninsula along the RW08 centreline, I see trees and terrain at my level. If I left the lateral protected area (0.2nm either side using RNP 0.10), I'd probably end up crashing. If I miss-set QNH (like I did in the video) I'd probably end up crashing.

 

Making a waypoint closer to the runway didn't seem to help. Anyone able to offer advice? I've added the bits I created for the PAJN.txt SIDSTAR file. Lateral and Vertical obstacle clearance isn't guaranteed as I don't know how to check easily.

 

FIX JN004 LATLON N 58 21.64000 W 135 19.87000

FIX JN003 LATLON N 58 21.49000 W 135 10.18000

FIX JN002 LATLON N 58 23.59000 W 134 50.68000

FIX JN001 LATLON N 58 22.80000 W 134 45.26000

 

APPROACH RNV08 FIX JN002 AT OR ABOVE 2100 FIX JN001 AT OR ABOVE 2100 RNW 08 FIX LEMNN FIX TRDWL FIX MARMN FIX GLAZZ HOLD AT FIX GLAZZ LEFT TURN INBOUNDCOURSE 125 ALT 7000 LEGTIME 1

TRANSITION CUSHI FIX CUSHI AT OR ABOVE 7000 FIX JN004 AT OR ABOVE 6400 FIX JN003 AT OR ABOVE 5200

 

At 1000' AGL, look the right of your nose and you'll see "The Cut," the lowpoint in the ridgeline.  You don't align with the runway until you fly through The Cut. Don't go over the ridgeline where you did.

 

Or that's what I'd do. Hypothetically.

[Copper. 78]

Thanks for the advice. I did build a procedure that took me over Coghlan thence thru the cut. I achieved that procedure, but I wanted to try tracking the extended centreline from atleast 5nm out to the threshold. Just so happens the cut is in the wrong spot. Who do I call to have it shifted?

 

I've tried moving the waypoint closer to the runway; that still gave me a 3.77 GP. I tried adding additional height to said waypoint - no affect.

 

I want to try a displaced threshold using the FMC/FMC Navdata before I abandon the exercise. Anyone know how to trick the FMC into thinking the threshold is further away?

[Stearmandriver 2,295]

Just curious... what makes you think the real world procedure tracks runway centerline for 5 miles, or even 1 mile? Matt gave you good advice. Theoretically. ;-)

As far as adjusting your GP angle, don't mess with the runway, just make the crossing altitude over the FAWP a hard altitude (not an at or above), and then adjust this crossing altitude until the box gives you a gp angle you like. A lot of this is trial/error and eyeballing. FWIW, building one of these procedures exactly as charted never works in the sim. You just gotta play around until it's close.

[Copper. 78]

Yeah, I know Matt is offering veeery sound advice ;D I'm emulating the GLS approach Boeing trialled using a Qantas frame during a bunch trials for ASA. The GLS took the Boeing over Mendenhall Hill clearing it by a few hundred feet. What I'm trying to create is a proc that reliably gets me over the hill with a bit more margin than I currently have.

 

Stuff of pure fiction as I said. I could (and have) probably replicate the RNP to RW08 near enough to the real one by flying thru the cut... But where's the fun in that?!

[Stearmandriver 2,295]

Haha got ya. I did not know that bit of history! But yeah, this stuff's fun, isn't it? A puzzle. You probably know that the airplane is limited to RNP approaches with a max steepness of 3.5 degrees. Not sure if that'll get you over the ridge, but much steeper and it's getting towards unstable approach territory... will be interesting to see if you can do it.

[Copper. 78]

It is very interesting. Queenstown (NZQN) has or had 3.77, so I assume that I can at least have that, but you're right... No matter how I do this, I'm not going to meet stable criteria at either 1000 or 500... The sink rate is probably the only factor making me unstable throughout the intermediate and final segment. Omitting the fact that I can't clear the land mass comfortably.

[Copper. 78]

https://youtu.be/njdgsA1p-k4

 

By amending the runway elevation in the Airport data, I squeezed an extra 60 feet clearance from the hill and created a pseudo displaced threshold. I've made a few other minor changes to fix locations to better align with the runway. I'm sure I can get more out of it.

 

Anyone able to tell me what the vertical minimum obstacle clearance might be for that hill using RNP 0.10, following the track that I made?

[ExNusquam 11]

OK, from TERPS for a Precision Approach; I'm just guessing that this is what the procedure would be built on, but it should get you pointed in the right direction. 

 

Obstacle Clearance Slope is S = 102/GPA, so in our case S=27:1

 

I'm just going to assume a reasonable TCH of 50 feet, for the purposes of calculation.

 

Height of the most restrictive OCS over the ridge is as follows: 

 

Z = (D- 200 + d)/S

 

Z= Height of OCS

S ="W" surface slope

d = d from paragraph 3.2.1 for GPI < 954' , 0 for GPI 954' or greater

Where D = the distance in feet from RWT

 

In our case, d=0, so Z=7500-200/27

Z= 270 feet, way lower than the ridge on centerline. The "cut" is only ~160' at it's highest.

 

In order to clear the ridge on centerline with the OCS, you'd need a 6.3°(!) glideslope, which would put you at 821'(!!) over the ridge. At that point, you'd be better off ditching the vertical profile and just going with a non-precision MDA with a VDP over the ridge. Non-precision approaches are controlling obstacle+250', which in this case would be lower at 700'. 

[Stearmandriver 2,295]

^^^ that assumes you're using the public TERPS criteria... Which is also the reason our friends in the double-breasted blazers can't just build their own approaches into JNU ;-). Still, in this case I agree that any reasonable clearance of the ridge on centerline privacy equals a descent rate that's more unstable than a lineup turn through the cut.

[Copper. 78]

 

OK, from TERPS for a Precision Approach; I'm just guessing that this is what the procedure would be built on, but it should get you pointed in the right direction.

 

Obstacle Clearance Slope is S = 102/GPA, so in our case S=27:1

 

I'm just going to assume a reasonable TCH of 50 feet, for the purposes of calculation.

 

Height of the most restrictive OCS over the ridge is as follows:

 

Z = (D- 200 + d)/S

 

Z= Height of OCS

S ="W" surface slope

d = d from paragraph 3.2.1 for GPI < 954' , 0 for GPI 954' or greater

Where D = the distance in feet from RWT

 

In our case, d=0, so Z=7500-200/27

Z= 270 feet, way lower than the ridge on centerline. The "cut" is only ~160' at it's highest.

 

In order to clear the ridge on centerline with the OCS, you'd need a 6.3°(!) glideslope, which would put you at 821'(!!) over the ridge. At that point, you'd be better off ditching the vertical profile and just going with a non-precision MDA with a VDP over the ridge. Non-precision approaches are controlling obstacle+250', which in this case would be lower at 700'.

you're an absolute legend! Thank you.